1

Managerial Ability, Analyst Recommendations, and Price Informativeness

Abstract

This paper examines whether managerial ability convey information about the

firm’s future earnings to the capital markets. Using the future earnings response coeffi-

cient (FERC) methodology, we find that the current stock returns of firms with

high-ability CEOs reflect more future earnings than does the stock returns of low-ability

firms. Our results are robust to controlling for potential omitted variables, loss versus

profit firms, and serial correlation of error terms. We further find that managerial abil-

ity has a positive impact on analyst recommendations for the firm. In addition, analyst

recommendations at least partially mediate the effects of managerial ability on FERC.

That is, analyst recommendations represent a mechanism through which managerial abil-

ity affects the extent to which stock returns reflects future earnings. Overall, this study

reveals the impact of managerial ability on analyst-based outcomes and current pricing

weight on expected earnings and calls attention to the construct of managerial ability as a

key intangible asset for the investor community.

Keywords: Managerial Ability; Future Earnings Response Coefficient (FERC); Analyst

Recommendations; Mediating Effect

2

1. Introduction

Managerial ability is clearly an important ingredient in the success of a business.

Previous research has focused on how managerial ability “matter” for corporate policies

and outcomes. For example, In Milbourn’s (2003) study, he focuses on the managerial

ability and measures it in terms of the number of press articles that cited the CEO. He

find that compensation contracts given to CEOs with a high ability (i.e., those with more

media-counts) exhibited greater equity instruments. Chemmanur and Paeglis (2005)

report that better quality management can affect the firm’s IPO characteristics. Some

studies have examined whether the reputational motivations of CEOs affect their deci-

sion-making behaviors such as investment and reporting decisions (Hirshleifer 1993;

Sridar 1994). Very few studies, however, have empirically examined whether managerial

ability sustains firm good performance and whether the expectation of superior perfor-

mance of highly ability CEOs, in fact, is delivered.

My focus on managerial ability is motivated by three considerations. First, mana-

gerial ability is one of the most important intangible assets that a firm has (Gaines-Ross,

2003). Second, it captures the dimension of managerial human capital (Francis, Huang,

Rajgopal and Zang 2008). Lastly, according to Burson-Marsteller’s survey in 1999, al-

most half of a firm’s reputation is based upon the image of its management team.

Thus this manager’s characteristic potentially can have a palpable impact on corporate

economic benefits. In fact, one of the reasons for the small number of empirical stud-

ies related to managerial ability is the difficulty of measuring the abilities of managers.

To overcome this concern, I measure managerial ability using four proxies which are

most often employed by previous studies: Press Citation, CEO awards from business

journals, manager-specific efficiency derived from data envelope analysis (DEA), and in-

dustry-adjusted firm performance (Milbourn 2003; Francis et al. 2004; Malmendier and

Tate 2005).

Empirical investigation of the economic benefits of managerial ability is important

because top executives with highly developed management skills often receive high pay.

The ability perspective advocates there exists economic benefits of managerial ability.

Most of previous studies have focused on the impact of managerial ability on the future

operating performance and yielded mixed evidence (Nanda et al. 1996; Malmendier and

Tate 2005). In this paper, I think that the important question is persistence of profits.

3

In fact, Gaines-Ross (2003) find that analysts recommend a company’s stock based on

CEO reputation because a CEO with a well-established reputation, believed to have high

ability, will sustain good performance or turn around poor performance. A board of

directors will give a highly regarded CEO more opportunities to make up for his or her

mistakes, because they believe that the CEO will reverse current poor performance in the

near term (Gaines-Ross 2003). Therefore, I expect that higher ability managers are able

to sustain stable performance over the long run (i.e., higher earnings persistence).

It is almost taken as an article of faith that high-ability managers help to firms’ in-

formation environment, thereby mitigating problems arising from information asym-

metry between the firm and the investors. Baik et al. (2011) report that managerial abil-

ity is positively related to the likelihood of voluntary management earnings forecasts.

This finding is consistent with highly reputed CEOs having incentives to keep the market

informed about their firms’ economic prospects. Moreover, managers with high abili-

ties may have higher ability to obtain more precise information on investment opportuni-

ties, make better investment choices and achieve successful project outcomes with greater

likelihood (Chemmanur, Paeglis, and Simonyan 2009). Successful corporate invest-

ments enhance firm future profits. Given that managerial ability provides useful infor-

mation about future profit streams, it helps investors to better predict future earnings.

Thus, more information about future earnings should be reflected in current stock prices.

Using future earnings response coefficient (FERC) as proxy for price informativeness, we

expect that it correlates positively with the level of managerial ability.

This research contributes to the literature in several ways. First, this study con-

tributes to the understanding of the economic benefits of managerial ability. A large

body of literature emphasizes the importance of managerial ability (Fama 1980; Gibbons

and Murphy 1992) and develops predictions about managerial behaviors (Hirshleifer

1993; Sridar 1994). However, scarce literature empirically investigates whether the abil-

ity of a CEO provides economic benefits to the firm. Second, although the value im-

pact of managerial ability is recognized (Agarwal, Taffler, and Brown 2007), the effects

of managerial ability on the ability of stock returns to reflect future earnings, however,

have been little studied. Our analysis provides fresh evidence about the information

content of managerial ability: the market understands and can better anticipate firms’

future earnings based on forward-looking information contained in managerial ability.

Third, given the scant evidence regarding the determinants of cross-sectional variation in

price informativeness with respect to future earnings, the inclusion of managerial ability

4

as an explanatory variable may improve our understanding of financial market behavior.

The rest of the paper is organized as follows. Section 2 provides a review of the

relevant literature and develops the hypotheses. In section 3, we describe our models,

variable measurements, sample selection and data sources.

2. Literature Review and Hypotheses

2.1 Importance of Managerial Ability

Managerial ability is viewed as a vital element in the capital market. Managers with

better ability can identify changes in their firm’s underlying economics. They convey

the intrinsic value of their firms more credibly to outsiders, thus reducing the infor-

mation asymmetry facing in the equity market. In addition, managerial ability is also a

factor in determining a firm’s financial and investment policies. The firms which have

talent manger access the equity market more easily, in turn to affect the firms’ financial

policies. Managers have force on the corporate decisions, such as investing and financ-

ing decision, executive compensation, and earnings quality. Managers with high reputa-

tions have incentives to protect their reputations by enhancing their firms’ performances.

Several studies have examined the impact of CEO characteristic on corporate deci-

sion. Chemmanura and Paeglisb (2005) suggest that IPOs firms with higher manage-

ment quality will be characterized by lower underpricing, greater institutional interest,

more reputable underwriters, and smaller underwriting expenses. Management reputa-

tion can reduce agency problems, managers intend to obtain funds and earn rents, and

firms must establish a reputation for generating profits and repaying these profits to in-

vestors. (Fama 1980, Narayanan 1985, Holmstrom and Ricart-Costa 1986, Holmstrom

1999) Chemmanur et al. (2009) indicate that better and more reputable managers may

be able to select better projects, characterized by a larger NPV for any given scale.

Taken together, these studies suggest that talent managers are able to cut down financing

costs, improve corporate governance and select the better investment project.

A substantial body of research is dedicated to understanding the determinants of

managerial ability and information environment. Trueman (1986) theorizes that man-

agers voluntarily issue earnings forecasts to signal their ability. Based on this theory,

Baik et al. (2010) find that firms with highly reputed CEOs are associated with lower

opacity, suggesting that CEOs act to protect their reputations by increasing the flow of

information to the market. They also views from an investor perspective and find that

5

firm value is increasing in CEO reputation, suggesting that investors perceive that highly

reputed CEOs improve the information environment. Baik et al. (2011) show that a

personal characteristic of the CEO is associated with the forecast accuracy increases in

CEO ability. The better managers may be able to recognize the firm’s quality, thus de-

clining the extent of information asymmetry between firm insiders and outsiders. This

means that managerial ability as a signal to help the investor can predict or determine

future profit

However, the extent literature focuses on the impact of managerial ability on firm

performance. Jensen and Fuller (2002) assert that CEO reputation is a major determi-

nant of the long-term success and survival of a firm. Chemmanura and Paeglisb (2005)

find that firms with better management quality have stronger its long-term post-IPO op-

erating performance and greater long-term stock returns. Agarwal et al. (2007) also

consider that good management enhances firm value: well managed firms have higher

profitability, are able to sustain superior operating performance for longer, and are re-

warded by higher market valuations. Switzer and Bourdon (2010) show several proxies

for management quality significantly affect operating performance of IPOs in Canada.

Above these papers only discuss the relation between managerial ability and current

firm’s performance. In this paper, we extend the line of research about the information

content of Managerial ability: whether better managerial ability help stock prices to re-

flect more future earnings given they convey forward-looking information.

2.2 Research on Future Earnings Response Coefficients (FERC)

Prior accounting literature has documented that stock prices reflect the expectations

of investors about future earnings, because the information can be aggregated with var-

ious public signals then reflected in the change in current stock price. Thus, the change

in current stock price captures the change in investors’ expectation for future earnings.

As the investors are better able to understand future earnings, more information about

future earnings will be reflected in current period stock returns. We follow the ap-

proach of Collins et al. (1994) and measure the extent to which current stock returns re-

flects future earnings by the future earnings response coefficient (FERC). Extent liter-

ature has used this approach to verify that FERCs can be enhanced by the better infor-

mation environment.

For example, Lundholm and Myers (2002) find that increased disclosure activity

“brings the forward” into current stock returns. Gelb and Zarowin (2002) consider that

6

the informativeness of firm disclosures influence FERCs. Ayers and Freeman (2003),

Piotroski ,Roulstone (2004) and Choi and Jung (2008) show that the informativeness of

stock returns for future earnings, measured as the FERC, increases with the number of

information intermediaries such as equity analysts and institutional investors. Choi et al.

(2011) find that the association between current returns and future earnings is increasing

the frequency and precision of management earnings per share forecasts.

In addition, researchers have begun to find more compelling evidence to verify the

association between corporate accounting policies and the future earnings response coef-

ficient. Ettredge et al. (2005) find the firm which adopts the disclosure rule SFAS

No.131 on segment reporting contains more information about its future earnings and

then reflected in higher FERC. Tucker and Zarowin (2006) find that managers use the

discretionary accruals to smooth income and garbles future accounting earnings infor-

mation. Oswald and Zarowin (2007) show that firms capitalize R&D has higher FERC

than expensing. Orpurt and Zang (2009) find that the direct method is valuable to in-

vestors with the higher FERCs. Hanlon et al. (2007) consider that firms initiate paying

improve the strength of relation between current returns and future earnings. Given

that managerial ability is a forward-looking indicator of future profits, this study add to

extent literature by investigating how managerial ability affect FERC.

2.3 Empirical Hypothesis

Since management ability has a critical impact on firm performance, this factor is an

important consideration in terms of investors’ firm valuation. Because managers may

have reputations for being skilled operators and cost cutters, the ability of a firm’s man-

agement can have a certifying effect on future profit streams. With regard to firm de-

velopment and performance, those firms with high ability management are also better

equipped to be more innovative (Staw and Epstein, 2000) and have higher internal gov-

ernance strength (Karuna 2010). Furthermore, highly ability managers have strong in-

centives to promote transparency by actively disclosing information about their firms’

economic prospects because the market values such information, thus reducing the in-

formation asymmetry that investors face when they predict future profits. Given that

managerial ability serves a forward-looking indicator of future profit streams, this metric

may help investors to better evaluate firm future earnings, which in turn reflected in cur-

rent stock returns. As a result, the association between current stock returns and future

earnings, FERC, should be higher for firms enjoying higher managerial ability. I for-

7

malize the second hypothesis as:

H1: The extent to which future earnings is reflected in current stock prices increases

with managerial ability.

The ability perspective found in the agency literature posits that the future firm

performance of more-skilled managers is likely to exceed that of less-skilled managers

(Gibbons and Murphy, 1992). Prior literatures have demonstrated hat high ability

management associated with higher profits next year than lower ranked management

(Wade et al. 2006; Nanda et al. 1996). This reasoning suggests that Managerial ability

can view as an indicator of more promising future firm profits. In addition, accounting

literatures suggest that financial analysts play an important role in providing investment

advice for individual and institutional investors. Extant research shows that analyst re-

leases stock recommendations to investors based on the prospects of firms’ future per-

formance. The better the prospects of firms’ future performance, the greater is the

likelihood that analyst will issue more favorable recommendations. Therefore, the bet-

ter managerial ability results in better prospects of firms’ future performance, positive

changes in managerial ability should lead analysts to recommend that investors hold or

buy a firm’s stock. I formalize the second hypothesis as follow:

H2: Analyst recommendations are positively associated with managerial ability.

Thus far, we have offered hypotheses on the impact of managerial ability on analyst

recommendations. Managerial ability affects analyst recommendations, which in turn

affect the FERC, it is reasonable to expect a “chained” relationship: from managerial

ability to the intermediate outcome of analyst recommendations and then to investors'

pricing weight on future earnings for equity valuation. It implies analysts are infor-

mation intermediaries between firms and investors, their recommendations likely act as

an informational channel through which the firm with better managerial ability and reach

investors. The more the firm enjoys favorable recommendations with highly ability

CEOs, the most likely the future earnings information impounded in managerial ability

is to pass through analyst recommendations and result in contributing to the FERC.

Given that analyst recommendations may convey the impact of managerial ability on the

FERC to investors. We suggest analyst recommendations s mechanisms for market re-

8

actions to the intangible asset of managerial ability. We establish the third testable hy-

pothesis is:

H3: Analyst recommendations at least partially mediate the associations between mana-

gerial ability and the FERC.

3. Methodology

3.1 Measures of Managerial Ability

It is difficult to measure the management quality directly and empirically. Conse-

quently, I employ four proxies to estimate the value of managerial ability. The first

measure we follow the approach of Milbourn (2003), Rajgopal et al. (2006), and Francis

et al. (2008) who use CEO press citations. A CEO who is perceived to be an expert is

more likely to be interviewed and cited in the newspaper and magazine. Thus, the press

citation measure of managerial ability likely reflects the market’s assessment of a CEO’s

perceived ability (Milbourn 2003). Similar to research employing press citations as a

proxy for managerial ability, we hand-collect CEOs’ press citations by searching news

articles from all publications in the Factiva database. We calculate the number of news

articles that mention the name of a CEO and the CEOs’ company. We assume that a

CEO develops his/her reputation over a number of years, so we sum the number of ci-

tations over the prior five years to measure ability.

For the second measure of managerial ability, we count the number of manager

awards from business journals as following Malmendier and Tate (2009). Various busi-

ness journals pick “CEOs of the Year” or “Best Managers” annual. We collected man-

ager award data from highly respected magazines such as Business Week, Financial World,

Forbes, Industry Week, Chief Executive, and Electronic Business. The winners of

manager awards from these business magazines are considered as a talent manager. We

set the dummy variable, equal to 1 if the manager was given any award for the last five

years or 0 otherwise.

We adopt the third measure is data envelope analysis (DEA). Demerjian et al. (2012)

use data envelope analysis (DEA) to derive a measure of CEO-specific talent. To re-

gress the firm-level measure on market share, size, the number of firm segments and

foreign operations, and firm fixed effects. The residual from this regression is the

measure of managerial ability. This measure of managerial efficiency can be considered

as managerial ability based on company performance. In validity checks of their meas-

9

ure, Demerjian et al. (2012) find that it is positively related to returns and CEO pay, and

that the persistence of earnings growth and sales growth is increasing in their measure of

managerial ability. Prior studies have linked earnings growth and sales growth with

managerial ability (e.g.Fee and Hadlock 2003), giving us added comfort that the DEA

measure is capturing some dimension of managerial ability.

The last proxy for managerial ability is the industry-adjusted ROA (IndAdjROA)

while the CEO has been at the helm of the firm (Milbourn, 2003). Under this indus-

try-adjusted measure, we posit that a positive value of the industry-adjusted measure

represents higher managerial ability. This approach provides a more objective estima-

tion, in that the market would not devalue a managerial ability due to poor operating

performance if most firms in the same industry also had such problems. Following the

approach in Milbourn (2003), IndAdjROA is calculated by using income before extraor-

dinary items scaled by average total assets for each firm and subtracting from it the aver-

age ROA for firms with the same two-digit SIC code for each firm-year. Then, we de-

leted observations if there were less than 10 firms within a two-digit SIC code for a given

year.

3.2 Primary Models and Empirical Predictions

The ability of stock prices to reflect future earnings can be measured by the FERC.

This paper fellows the approach of Collins et al. (1994), Lundholm and Myers (2002),

and Tucker and Zarowin (2006). To reduce the measurement error problem in using

realized earnings for expected earnings CKSS include future returns. Based on the prior

studies, we recognize lag stock returns to measure value creation.

(1) 34332110 tttttt RbXbXbXbbR

In regression (1), Rt is current annual returns for Year t. Xt-1 and Xt are the earn-

ings per share (EPS) for Year t-1 and t, respectively, and Xt3 represents the realized future

earnings aggregate over three years with annual compounding. All the EPS variables

are the basic EPS excluding extraordinary items, adjusted for stock splits and stock divi-

dends, and according to Christie (1987), deflated by the stock price at the beginning of

Year t. Rt3 is common stock return for three year period starting from three months

after t fiscal year-end with annual compounding. In regression (1), the coefficient b3 is

FERC, and is expected to be positive. The coefficient on past earnings (b1) is predicted

10

to be negative, the ERC (b2) is predicted to be positive, and coefficient on future returns

(b4) is predicted to be negative.

The purpose of this paper is to examine whether better managerial ability help

stock prices to reflect more future earnings given they convey forward-looking infor-

mation. Thus, we expand the above regression by adding the managerial ability, denot-

ed as ABILITYt and we interact ABILITYt with earnings variables. The empirical mod-

el is as follows:

(2) 39387

16534332110

ttttttt

tttttttt

RABILITYbXABILITYbXABILITYb

XABILITYbABILITYbRbXbXbXbbR

In regression (2), we are interest in b8 if the managerial ability is to convey infor-

mation about future earnings, then the coefficient on 3tt XABILITY should be positive.

If the managerial ability contains less forward-looking information, then the coefficient

on 3tt XABILITY should be negative. We have no prediction for the coefficients on

tABILITY , 1 tt XABILITY , tt XABILITY and 3tt RABILITY .

In order to prevent some other factors may make the stock price impound more in-

formation about future earnings, we control for firm size (SIZE), growth opportunities

(BM), future earnings variability (EARNSTD), and analyst following (NANAL). Firm

size is measured as the market value of common equity at the beginning of Year t.

Growth opportunities are measured by the book-to-market ratio at the beginning of Year

t. For future earnings variability, we measure by the standard deviation of EPS from

Year t+1 to Year t+3, and then deflated by the stock price at the beginning of Year t.

Analyst following is measured as the average number of analysts’ forecasts in the

monthly consensus, gather from I/B/E/S during Year t.

Our H2 is to ascertain whether security analysts are influenced by managerial ability.

If analysts recognize that managerial ability could indicate better earnings prospectus, we

should observe a positive association between consensus recommendations and managerial

ability. We test for such an effect using multivariate estimations in which the dependent

variable equals the level of consensus recommendation, and the explanatory variables of in-

terest is managerial ability proxies (ABILITY). We include several factors that are known

to influence recommendations (e.g. Jegadeesh et al., 2004). These factors include the mar-

ket-adjusted returns over the past twelve months (RETP) to capture momentum effects, av-

erage daily turnover (TURN), sales growth (SG), earnings / price (EP), and book / market

11

(BM) to capture contrarian effects. We additionally include firms size (SIZE), analyst fore-

cast revisions (FREV), standardized unexpected earnings (SUE), total accruals (TA), and

capital expenditures (CAPEX). Specifically, the model is specified as follows:

(3) 11109876

543210

ttttttt

tttttt

EPbBMbCAPEbTAbSGbSUEb

FREVbSIZEbTURNbRETPbABILITYbbREC

where analyst recommendations (REC) are the expert advice from financial analysts to

investors before the actual earnings announcements at the end of the year. The meas-

ure of analyst recommendation is a Likert scale which ranges from “1”=

sell-recommendation for the stocks to “5”= strong buy-recommendation for the stocks

(Bryan and Tiras 2007; Womack 1996). All of independent variables are defined in the

Table 1. We focus on the coefficient b1. We predict b1 to be positive indicating that

managerial ability improves analyst recommendations.

To test the mediating role of analyst recommendations in managerial ability’s possi-

ble impact on the FERC, we use the approach of Baron and Kenny (1986). We extend

regression (2) by including Analyst recommendations (REC) and its interactions with

independent variables as follows:

(4) * 3111039

387

16534332110

tttttt

tttt

tttttttt

XRECbRECbRABILITYb

XABILITYbXABILITYb

XABILITYbABILITYbRbXbXbXbbR

If the impact of managerial ability on the FERC is decreased after the inclusion of

analyst activities, that suggests evidence for the partial mediation role of analyst-based

processes.

3.3 Data and Sample Selection

We gather data from CRSP and Compustat databases. The data period is start from

1990 to 2007. The sample ends in 2007 because we require three years subsequent

earnings and returns in formation. We exclude regulated utilities (SIC 40-49) and finan-

cial institutions (SIC 60-69) from the initial sample. To minimize the effect of outliners,

we follow Tucker and Zarowin (2006) and delete observations that are in the top or bot-

tom 0.5 percent of the distributions of past, current, and future three years’ earnings,

and of current and future three years’ returns.

12

4. Empirical Results

4.1 Variables and Descriptive Statistics

Panel A of Table 2 reports descriptive statistics for the variables used in the FERC

regressions. Current annual returns (Rt) have a mean (median) of 11.6 percent (7.2

percent), and earnings per share deflated by the beginning svtock price have the mean

(median) of 4.3 percent (5.0 percent). We use four proxies to measure the managerial

ability. The mean number of press citations is 2.5 citations and about 10% of CEOs

received an award in the past 5 years. Note that, the ability variable, as measured by the

total number of media citations, is highly skewed. We adopt two ways to fix this in later

analysis: first, we take the logarithm of the citation plus 1; second, we define the dummy

variables for CEOs of low and high ability. Both measures give economically similar

results in our tests. So we only report the results using the logarithm of (citiation+1).

Managerial ability measured by DEA score has a mean and median close to zero, by con-

struction, as this is a residual from firm efficiency model. The standard deviations of

future earnings (EARNSTD) have a mean (median) of 3.8 percent (1.8 percent).

Unreported correlation analyses show that current stock returns, Rt, are positively

correlated with current and future earnings (Xt and Xt3), as expected. Returns are nega-

tively correlated with past earnings, Xt-1, in line with the mean-reverting nature of earn-

ings. As expected, the future returns variable, Rt3 are positively correlated with future

earnings, Xt3. However, one concern is the statistically significant negative correlation

between current returns, Rt, and future returns, Rt3 (Pearson correlation=-0.182 and

Spearman correlation=-0.093). As a result, future returns may influence our regression

results beyond their role as a measurement error proxy. Indeed, Orpurt and Zang (2009)

also show a significant correlation between these variables in their Panel D of Table 2.

Both Pearson and Spearman correlations show that four proxies of managerial ability,

ABILITYt is positively correlated with future earnings figures. Besides, multicollinearity

is not a serious concern as the variance inflation factor (VIF) is under 10 for all our re-

gression specifications (Belsley et al. 1980).

Panel B of Table 2 presents descriptive statistics for the analyst recommendation model.

The mean (median) sample firm receives a three-tier consensus recommendation of 2.68

(2.66), which is between sell and hold. The sample is tilted toward large firms as the aver-

age company in our sample has market capitalization of $1.4 billion. The average sample

13

firm has a book-to-market ratio of 0.449, sales growth of about 1.15% compared to the pri-

or four quarters, capital expenditures equaling about 6.5% of assets, and total accruals

equaling about 4.9% of assets.

Unreported correlation analyses show that analyst recommendations have a correlation

of 0.28 – 0.09 with the managerial ability measures. Analyst recommendations are also

positively correlated with past returns (0.10 – 0.14), standardized unexpected earnings (0.16),

and sales growth (0.19). Recommendations are negatively correlated with the firm’s

book-to-market ratio (–0.22). All of these results are consistent with Jegadeesh et al. (2004).

Our managerial ability measures generally have low correlations with firm characteristics,

although CEO award is positively correlated with the natural logarithm of market capitaliza-

tion (0.39) and the sales growth (0.27).

4.2 The Effect of Managerial Ability on the FERC (H1)

In this section, we use OLS regression analyses to test whether firms with higher

managerial ability have greater FERC than firms with lower ability. Standard errors for

all regressions are corrected for heteroskedasticity and within-firm clusters. To com-

pare with previous studies, in Panel A of Table 3 we first present the results of the base-

line FERC model (Eq. (1)). In the first column, as previously predicted, both the coef-

ficient on Xt3 (FERC) and the coefficient on Xt, are significantly positive (0.639 and 0.717,

respectively), and the coefficient on Xt-1 is significantly negative (-0.698). The coeffi-

cient on Rt,1-3 is also significantly negative (-0.086), which confirms the successful role of

the instrumental variable. The adjusted R2 of the baseline model is 12.2 percent, which

is higher than that of the traditional ERC model (7.64 percent). The significant im-

provement in explanatory power is consistent with the findings of Collins et al. (1994).

Our primary objective is to measure the impact of managerial ability on the FERCs.

We estimate four separate models, one for each measure of managerial ability. The re-

sults of the regression analysis are shown in right four columns of Panel A. After in-

cluding the interaction between Abilityt and the other variables in the benchmark FERC

model, the coefficients on Xt and Xt3 are significantly positive and the coefficients on Xt-1

and Rt3 are significantly negative. The focus of this table is on the regression coefficient

of ABILITYt *Xt3, corresponding to b8 in Eq. (2). The results show that all of four abil-

ity interaction variables exhibit positive and significant coefficients (b8 ranging from 0.008

to 1.048 at the 5% level or even lower); it suggests that the extent to which current re-

turns reflect the future earnings increases with managerial ability. The results strongly

support H1 and suggest that managerial ability conveys valuable information about fu-

14

ture earnings capacity that investors can use to place a greater current pricing weight on

future expected earnings.

We then perform analyses to control for potential correlated omitted variables.

Based on the work of Lundholm and Myers (2002) and Tucker and Zarowin (2006), firm

size, negative earnings, firm growth, earnings volatility, and analyst following have all

been shown to be significantly related to the FERC. Thus, we estimate the model simi-

lar to Eq. (2) above but with the addition of these explanatory variables, each separately

interacted with the future earnings variable as well as included simultaneously.

Panel B of Table 3 shows the estimation results for the models adding the full con-

trol variable. After controlling for these factors, the coefficients on ABILITYt *Xt3 re-

main statistically positive, strongly supporting our hypothesis that managerial ability pro-

vides valuable information and thus improves market’s ability to impound future earn-

ings into current returns. Besides, the coefficients on Xt3*SIZEt are positive, suggesting

that firm size is associated with an increase in the ability of stock prices to reflect future

earnings. In terms of BMt variable, growth may take time to be materialized and recog-

nized in earnings while the return reflects expectation for future growth immediately.

Thus, growth firms have a higher FERC than mature firms. The coefficients on

Xt3*LOSSt and Xt3*EARNSTDt are statistically negative, suggesting that the extent to

which stock returns reflect future earnings are smaller for firms reporting losses and for

firms with more volatile earnings stream. Consistent with the findings of Tucker and

Zarowin (2006), we find that the association between current returns and future earnings

is decreasing in the number of analyst following. Overall, the findings support that

managerial ability is associated with stock prices that reflect more information about fu-

ture earnings, attesting to the important role of management quality in providing relevant

information about future earnings.

4.3 Managerial Ability and Analyst Recommendations (H2)

Panel A of Table 4 presents regression results for test of H2. The dependent varia-

ble equals the level of analysts’ consensus recommendation. The first through fourth

columns provide the estimates from four models, one for each measure of managerial

ability. Standard errors are clustered both along the year and firm dimensions. The

coefficient estimates of ABILITYt are positive and highly significant, indicating that

higher level of managerial ability is associated with more favorable consensus recom-

mendations, after controlling for the known determinants of recommendations. The

15

results strongly support H2 and suggest that analysts have a favorable view of managerial

ability, possibly because management quality contributes with the most productive and

efficient way of enabled resources within the organization, thereby achieving long term

corporate value creation. This effect is significant in economic terms. When manage-

rial ability is measured by DEA score, a one standard deviation increase in ABILITY in-

creases the consensus recommendation by 0.02, which is similar to the impact of

well-known determinants. For example, a one standard deviation increase in EP or SG

increases the consensus recommendation by 0.01 or 0.06, respectively. Given that ana-

lyst recommendations tend to concentrate above 2.5, these small incremental effects are

potentially large in economic importance.

Analysts issue more favorable recommendations to stocks with positive price mo-

mentum, higher sales growth, lower turnover, lower book/market ratios, and higher

earnings/price ratios. Firms with higher earnings surprises, more positive analyst earn-

ings forecast revisions, greater total accruals, and more capital expenditures are associated

with higher recommendations. These estimates are generally consistent with prior liter-

ature.

4.4 The Mediating Role of Managerial ability (H3)

H3 predict that analyst recommendations at least partially mediate the association

between managerial ability and the FERC. According to Baron and Kenny (1986), to

establish mediation, managerial ability must affect analyst recommendations, and analyst

recommendations must affect the FERC. As we found previously, managerial ability

affects analyst recommendations. In additional, the results of Table 3 suggest that

managerial ability has effect on the FERC. As we report in Table 5, because inclusion

of analyst recommendations in the model reduces the strength of the effects of manage-

rial ability on the FERC (from 0.07, p=0.094, to 0.563, p=0.057), only marginally signifi-

cant), the results support a partial mediating role of analyst recommendations. In addi-

tion, the inclusion of mediating effects of recommendations significantly improves the

fit of the FERC models, as Table 5 shows. Specifically, adding the direct and mediating

effects of recommendations leads to an incremental R-square ranging from 0.06 to 0.08

in various model specifications, thus explaining significantly more variance of firm stock

returns.

Finally, we conduct Sobel’s (1982) test for mediation to assess whether the indirect

mediation effects are statistically significant. The standard Sobel test model is z = ab /

16

, where a and are coefficient and standard error, respective-

ly, for the impact of the independent variable on the mediator and b and are coeffi-

cient and standard error, respectively, for the impact of the mediator on the dependent

variable. We find that the Sobel test results are consistently significant (smallest z-value

= 2.32, p<0.05) for all indirect mediation effects. Thus, managerial ability’s indirect ef-

fects through the mediating role of recommendations are mostly significant, supporting

the prediction of H3.

5. Conclusion

In this paper, we examine whether managerial ability are informative about the

firm’s future earnings. If information regarding future earnings is revealed by manage-

rial ability, the stock price should reflect more of future earnings news for firms with

higher quality management. Our empirical analyses reveal that the current stock returns

of firms with higher managerial ability reflect more future earnings than does the returns

of lower-ability firms. Our results remain after controlling for potential omitted varia-

bles. We conclude that the market attaches value to managerial ability, because the abil-

ity enables investors better anticipate a firm’s future profitability. We further investigate

whether the stock recommendations of security analysts are influenced by firm manage-

rial ability. If security analysts identify managerial ability as a factor that contributes to

future earnings streams, they will issue more favorable recommendations to stocks with

quality management. We empirically find that managerial ability has a positive impact

on analyst recommendations for the firm. In addition, analyst recommendations at

least partially mediate the effects of managerial ability on FERC. That is, analyst rec-

ommendations represent a conduit through which managerial ability influences investors’

pricing weights of future earnings component in current stock returns. Given sophisti-

cated analysts pay attention to management quality, market investors should have good

reason to listen and follow. As such, the findings of this study also have practical im-

plications.

17

References

Agarwal, V., Taffler, R., & Brown, M. (2007). Is management quality value relevant?

Working paper, University of Edinburgh.

Ayers, B. C., & Freeman, R. N. (2003). Evidence that analyst following and

institutional ownership accelerate the pricing of future earnings. Review

of Accounting Studies, 8, 47–67.

Baik, B., Brockman, P., Farber, D. B., & Lee, S. (2010). CEO reputation and

corporate opacity. Working paper, Seoul National University.

Baik, B., Farber, D. B., & Lee, S. S. (2011). CEO ability and management earnings

forecasts. Contemporary Accounting Research, 28, 1645–1668.

Baron, R. M., & Kenny, D. A. . (1986). The moderator-mediator variable distinction

in social psychological research: Conceptual, strategic, and statistical

considerations. Journal of Personality and Social Psychology, 51,

1173-1182.

Belsley, D. A., Kuh, E., & Welsch, R. E. . (1980). Regression diagnostics:

Identifyinginfluentialdata and sources of collinearity. Wiley, New York.

Bryan, D. M., & Tiras, S. L. (2007). The influence of forecast dispersion on the

incremental explanatory power of earnings, book value, and analyst

forecasts on market prices. Accounting Review, 82, 651-677.

Chemmanur, T. J., Paeglis, I., & Simonyan, K. (2009). Management quality, financial

and investment policies, and asymmetric information. Journal of Financial

& Quantitative Analysis, 44, 1045-1079.

Chemmanura, T. J., & Paeglis, I. (2005). Management quality, certification, and

initial public offering. Journal of Financial Economics, 76, 331-368.

Choi, B., & Jung, K. (2008). Analyst following, institutional investors and pricing of

future earnings: Evidence from Korea. Journal of International Financial

Management & Accounting, 19, 261-286.

Choi, J.-H., Myers, L. A., Zang, Y., & Ziebart, D. A. (2011). Do management EPS

forecasts allow returns to reflect future earnings? Implications for the

continuation of management's quarterly earnings guidance. Review of

Accounting Studies, 16, 143-182.

Christie, A. (1987). On cross-sectional analysis in accounting research. Journal of

Accounting and Economics, 9, 231–258.

Collins, D. W., Kothari, S. P., Shanken, J., & Sloan, R. G. (1994). Lack of timelines

18

and noise as explanations for the low contemporaneous return-earnings

association. Journal of Accounting & Economics, 18, 289-324.

Demerjian, P., Lev, B., McVay, S. (2012). Quantifying managerial ability: A new

measure and validity tests. Management Science, Forthcoming, 1-20.

Ettredge, M. L., Kwon, S. Y., Smith, D. B., & Zarowin, P. A. (2005). The impact of

SFAS No. 131 business segment data on the market’s ability to anticipate

future earnings. Accounting Review, 80, 773–804.

Fama, E. F. (1980). Agency problems and the theory of the firm. Journal of

Political Economy, 88, 288-307.

Fee, T., & Hadlock, C. (2003). Raids, rewards, and reputations in the market for

managerial talent. Review of Financial Studies, 16, 1315-1357.

Francis, J., Huang, A. H., Rajgopal, S., & Zang, A. Y. (2008). CEO reputation and

earnings quality. Contemporary Accounting Research, 25, 109-147.

Francis, J., LaFond, R., Olsson, P., & Schipper, K. (2004). Costs of equity and

earnings attributes. Accounting Review, 79, 967 – 1010.

Gaines-Ross, L. (2003). CEO Capital: A Guide to Building CEO Reputation and

Company Success, New York: Wiley.

Gelb, D. S., & Zarowin, P. A. (2002). Corporate disclosure policy and the

informativeness of stock returns. Review of Accounting Studies, 7, 33–52.

Gibbons, R., & Murphy, K. J. (1992). Optimal incentive contracts in the presence of

career concerns: Theory and evidence. Journal of Political Economy, 100,

468.

Hanlon, M., Myers, J., & Shevlin, T. (2007). Are dividends informative about future

earnings? Working paper, University of Arkansas.

Hirshleifer, D. (1993). Managerial reputation and corporate investment decisions.

The Journal of the Financial Management Association, 22, 145.

Holmström, B. (1999). Managerial incentive problems: A dynamic perspective.

Review of Economic Studies, 66, 169-182.

Holmstrom, B., & Costa, J. R. I. (1986). Managerial incentives and capital

management. Quarterly Journal of Economics, 101, 835-860.

Jegadeesh, N., Kim, J., Krische, S. D., & Lee, C. M. C. (2004). Analyzing the Analysts:

When Do Recommendations Add Value? Journal of Finance, 59, 1083-1124.

Jensen, C. M., & Fuller, J. (2002). What's a director to do? negotiations,

organization and markets research papers. Harvard NOM Research Paper

No. 02-38.

Karuna, C. (2010). CEO reputation and internal corporate governance. Working

paper, University of Houston.

Leverty, J., & Grace, M. (2005). Dupes or incompetents? An examination of

19

management's impact on property-liability insurer distress. Working

paper, Georgia State University.

Lundholm, R., & Myers, L. A. (2002). Bringing the future forward: The effect of

disclosure on the returns-earnings relation. Journal of Accounting

Research, 40, 809-839.

Malmendier, U., & Tate, G. (2005). CEO overconfidence and corporate investment.

Journal of Finance, 60, 2661-2700.

Malmendier, U., & Tate, G. (2009). Superstar CEOs. Quarterly Journal of Economics,

124, 1593-1638.

Milbourn, T. T. (2003). CEO reputation and stock-based compensation. Journal of

Financial Economics, 68, 233.

Nanda, S., Schneeweis, T., & Eneroth, K. (1996). Corporate performance and firm

perception: The British experience. European Financial Management, 2,

197-221.

Narayanan, M. P. (1985). Managerial incentives for short-term results. Journal of

Finance, 40, 1469-1484.

Orpurt, S. F., & Zang, Y. (2009). Do direct cash flow disclosures help predict future

operating cash flows and earnings? Accounting Review, 84, 893–935.

Oswald, D. R., & Zarowin, P. (2007). Capitalization of R&D and the

informativeness of stock prices. European Accounting Review, 16,

703–726.

Piotroski, J. D., & Roulstone, D. T. (2004). The influence of analysts, institutional

investors, and insiders on the incorporation of market, industry, and

firm-specific information into stock prices. Accounting Review, 79,

1119-1151.

Rajgopal, S., Shevlin, T., & Zamora, V. (2006). CEOs' outside employment

opportunities and the lack of relative performance evaluation in

compensation contracts. Journal of Finance, 61, 1813-1844.

Sobel, M. E. (1982). Asymptotic Intervals for Indirect Effects in Structural

Equations Models. in Sociological Methodology, S. Leinhart, ed. San

Francisco: Jossey-Bass, 290–312.

Sridhar, S. S. (1994). Managerial reputation and internal reporting. Accounting

Review, 69, 343-363.

Staw, B. & Epstein, L. (2000). What the bandwagons bring: Effects of management

techniques on corporate performance, reputation and CEO pay. Adminis-

trative Science Quarterly, 45, 523 – 556.

Switzer, L. N., & Bourdon, J.-F. (2010). Management quality and operating

performance: Evidence for Canadian IPOs. working paper,Concordia

20

University.

Trueman, B. (1986). Why do managers voluntarily release earnings forecasts?

Journal of Accounting and Economics, 8, 53-71.

Tucker, J. W., & Zarowin, P. A. (2006). Does income smoothing improve earnings

informativeness? The Accounting Review, 81, 251–270.

Wade, J. B., Porac, J. F., Pollock, T. G., & Graffin, S. D. (2006). The burden of celeb-

rity: The impact of CEO certification contests on CEO pay and perfor-

mance. Academy of Management Journal, 49, 643-660.

Womack, K. L. (1996). Do Brokerage Analysts' Recommendations Have Invest-

ment Value? Journal of Finance, 51, 137-167.

21

TABLE 1 Variable Measurements

Panel A: FERC model

Variables Measurements

Rt The buy-and-hold annual returns during the 12-month period starting three months following the firm's t-1 fiscal year-end.

Xt-1 Earnings per share for Fiscal Year t-1, deflated by the stock price at the beginning of Fiscal Year t.

Xt Earnings per share for Fiscal Year t, deflated by the stock price at the beginning of Fiscal Year t.

Xt3 The sum of earnings per share for Fiscal Years t+1 through t+3, de-flated by the stock price at the beginning of Fiscal Year t.

Rt3 The buy-and-hold stock return for Fiscal Years t+1 through t+3 starting three months following the firm's t fiscal year-end.

SIZEt Market capitalization ($millions) at the beginning of Fiscal Year t. We use log transformation in correlation and regression analyses.

BMt Book value of equity divided by market value of common shares out-standing at the beginning of Fiscal Year t.

EARNSTDt The standard deviation of earnings per share for Fiscal Years t+1 to t+3, deflated by the stock price at the beginning of Fiscal Year t.

NANALt The number of analysts following in the latest month prior to earnings announcement for Fiscal Year t. We use log transformation in correla-tion and regression analyses.

Panel B: Analyst recommendation model

Variables Measurements

RECt The mean of analyst recommendations made or reviewed at least two days but no more than a year prior to the calendar year’s end. If an analyst had more than one recommendation in this window, the newest one is taken. The recommendations are coded on a scale from 1 (strong sell) to 5 (strong buy). REC is a continuous variable (1, 5).

SIZEt Natural logarithm of market capitalization at the end of each calendar year.

RETPt Market adjusted buy-end-hold return over the month (-1,-12) prior to the calendar year’s end. The market return is the value-weighted return from CRSP.

FREVt The sum of price-scaled mean forecast revisions (taken from the IBES summary statistics) using months (-1, -12) relative to the end of each cal-endar year.

SUEt SUE: Standardized unexpected earnings. (EPSq-EPSq-4)/ stdq, where EPS is earnings per share excluding extraordinary items from the most recent quarterly earnings announcement prior to the end of the calendar year, and the denominator is the standard deviation of unexpected earnings (EPSt-EPSt-4) from quarters t=q-7 to t=q.

EPt Sum of the past four quarters earnings per share divided by the stock price at the end of the most recent fiscal year with an earnings announcement preceding the current calendar year end.

22

BMt Book value of asset scaled by the sum of book value of assets, negative one times long-term debt (including current part), and market value of common stock at the end of the most recent fiscal year with an earnings announcement preceding the current calendar year end.

TAt (total accrualst-total accrualst-1)/(total assetst), where total accruals is de-fined as change in current assets minus change in cash, minus change in current liabilities, plus change in long-term debt, minus change in deferred taxes, minus depreciation. Changes are from four quarters earlier. t refers to the year of the most recent fiscal year with an earnings announcement preceding the current calendar year end.

CAPEXt Sum of capital expenditures from the most recent four quarters scaled by total assets

SGt Sales growth, defined as the sum of sales in quarters 0,-3 scaled by the sum of sales in quarters -4,-7, where quarter zero refers to the end of the most recent fiscal quarter with an earnings announcement preceding the current calendar year end.

TURNt Percentile rank of the daily average turnover per month, averaged over the past six months, percentiles calculated by exchange, where turnover is de-fined as share volume over shares outstanding.

23

TABLE 2 Sample Statistics

Panel A：Descriptive Statistics

Variable N Mean Median Std. Dev. Q1 Q3

Rt 10,286 0.116 0.072 0.427 -0.142 0.305 Xt-1 10,286 0.041 0.048 0.072 0.029 0.066 Xt 10,286 0.043 0.050 0.069 0.028 0.070 Xt3 10,286 0.129 0.138 0.197 0.063 0.215 Rt3 10,286 0.295 0.111 0.886 -0.240 0.577 DJHITSt 10,286 2.489 0.000 9.919 0.000 2.000 AWARDt 10,286 0.096 0.000 0.295 0.000 0.000 MGTEFFt 7,044 -0.040 -0.033 0.128 -0.111 0.017 ADJROAt 10,226 0.058 0.039 0.091 0.005 0.092 SIZEt 10,286 7.555 7.422 1.446 6.500 8.461 LOSSt 10,286 0.104 0.000 0.305 0.000 0.000 BMt 10,286 0.449 0.399 0.285 0.254 0.579 EARNSTDt 10,286 0.038 0.018 0.059 0.009 0.041 NANALt 10,286 2.403 2.398 0.585 1.946 2.833

Panel B：Descriptive Statistics ─ Consensus Recommendation Level

RECt 10,286 2.678 2.667 0.567 2.333 3.000 RETPt 10,286 0.116 0.072 0.427 -0.142 0.305 TURNt 10,286 51.702 53.000 28.427 27.000 76.000 SIZEt 10,286 7.555 7.422 1.446 6.499 8.461 FREVt 10,286 0.003 0.004 0.034 -0.002 0.010 SUEt 10,286 -0.425 0.278 28.783 -0.799 1.187 SGt 10,286 1.147 1.107 0.219 1.040 1.206 TAt 10,286 -0.049 -0.046 0.080 -0.081 -0.015 CAPEXt 10,286 0.065 0.047 0.065 0.024 0.083 BMt 10,286 0.449 0.399 0.285 0.254 0.579 EPt 10,286 0.040 0.046 0.079 0.028 0.064

24

TABLE 3 Regressions of FERC Models: Analyzing the Effect of Mangerial Ability

Panel A: Main Tests BaselineModel Managerial ability (ABILITY) proxies =

Not Included Press

Citation Award DEA Score IndAdjROA

Variables

Coefficient estimate

Coefficient estimate

Coefficient estimate

Coefficient estimate

Coefficient estimate

Intercept 0.053*** 0.059*** 0.069*** 0.058*** 0.055***

(0.000) (0.000) (0.000) (0.000) (0.000) Xt-1 -0.698*** -0.700*** -0.848*** -0.766*** -0.859*** (0.000) (0.000) (0.000) (0.000) (0.000) Xt 0.717*** 0.691*** 0.756*** 0.833*** 0.754*** (0.000) (0.000) (0.000) (0.000) (0.000) Xt3 0.639*** 0.627*** 0.600*** 0.725*** 0.614*** (0.000) (0.000) (0.000) (0.000) (0.000) Rt3 -0.086*** -0.088*** -0.094*** -0.087*** -0.097*** (0.000) (0.000) (0.000) (0.000) (0.000) ABILITYt -0.003*** -0.098*** 0.006 -0.048 (0.000) (0.000) (0.900) (0.335) ABILITYt * Xt-1 0.007 0.096 -0.906 -0.917*** (0.498) (0.780) (0.149) (0.006) ABILITYt * Xt 0.004 -0.181 1.158* 1.903*** (0.564) (0.554) (0.078) (0.008) ABILITYt * Xt3 0.008** 0.575*** 0.534** 1.048*** (0.013) (0.000) (0.028) (0.000) ABILITYt * Rt3 0.001** 0.038** 0.090** 0.183*** (0.016) (0.030) (0.033) (0.000) Adjusted R2 0.122 0.124 0.123 0.126 0.125 Observations 10,286 10,286 10,286 7,044 10,226

25

Panel B: Including Potentially Correlated Omitted Variables Adding a single control variable

Managerial ability (ABILITY) proxies =

Press Citation Award DEAScore IndAdjROA

Variables

Coefficient estimate

Coefficient estimate

Coefficient estimate

Coefficient estimate

Intercept 0.074** (0.020)

0.132***

(0.000)

-0.010 (0.787)

0.070**

(0.026)

Xt-1 -0.623*** (0.000)

-0.750*** (0.000)

-0.703*** (0.000)

-0.839*** (0.000)

Xt 0.434*** (0.000)

0.565***

(0.000)

0.551*** (0.000)

0.571***

(0.000)

Xt3 1.258*** (0.000)

1.026***

(0.000)

1.060*** (0.000)

0.804***

(0.000)

Rt3 -0.103*** (0.000)

-0.112*** (0.000)

-0.102*** (0.000)

-0.113*** (0.000)

ABILITYt -0.002** (0.015)

-0.055** (0.022)

0.036

(0.434)

0.194*** (0.000)

ABILITYt *Xt-1 0.009 (0.354)

0.064

(0.849)

-1.194* (0.051)

-0.686** (0.036)

ABILITYt *Xt 0.007 (0.330)

-0.141 (0.636)

0.746

(0.244)

2.179*** (0.002)

ABILITYt *Xt3 0.010*** (0.002)

0.420***

(0.002)

0.632*** (0.009)

0.766**

(0.010)

ABILITYt * Rt3 0.001* (0.075)

0.047***

(0.006)

0.084** (0.041)

0.197***

(0.000)

SIZEt -0.051*** (0.000)

-0.061*** (0.000)

-0.053*** (0.000)

-0.059*** (0.000)

SIZEt * Xt3 0.002* (0.916)

0.067***

(0.001)

0.058** (0.047)

0.078***

(0.000)

LOSSt -0.048*** (0.004)

-0.032* (0.064)

-0.067*** (0.001)

-0.043** (0.013)

LOSSt * Xt3 -0.199*** (0.000)

-0.209*** (0.000)

-0.128* (0.061)

-0.113** (0.036)

BMt 0.069*** (0.000)

0.081***

(0.000)

0.133*** (0.000)

0.129***

(0.000)

BMt * Xt3 -0.129** (0.019)

-0.162*** (0.002)

-0.087 (0.241)

-0.140** (0.016)

EARNSTDt 0.700*** (0.000)

0.145***

(0.000)

0.758*** (0.000)

0.707***

(0.000)

EARNSTDt * Xt3 -0.969*** (0.000)

-0.709*** (0.000)

-1.102*** (0.000)

-0.830*** (0.000)

NANALt 0.122*** (0.000)

0.128***

(0.000)

0.151*** (0.000)

0.130***

(0.000)

NANALt * Xt3 -0.133*** (0.006)

-0.234*** (0.000)

-0.164** (0.012)

-0.185*** (0.000)

Adjusted R2 0.165 0.168 0.178 0.170 Observations 10,286 10,286 7,044 10,226

26

TABLE 4 Regressions of Consensus Recommendation Level (“REC”)

Panel A: Main Tests

Managerial ability (ABILITY) proxies =

Press Citation Award DEA Score IndAdjROA

Variables

Coefficient estimate

Coefficient estimate

Coefficient estimate

Coefficient estimate

Intercept 2.889*** 2.864*** 2.846*** 2.831***

(0.000) (0.000) (0.000) (0.000)

ABILITYt 0.002*** 0.080*** 0.145*** 0.128*

(0.001) (0.000) (0.005) (0.052)

RETPt 0.233*** 0.226*** 0.225*** 0.232***

(0.000) (0.000) (0.000) (0.000)

TURNt -0.001*** -0.001*** -0.002*** -0.001***

(0.000) (0.000) (0.000) (0.000)

SIZEt -0.044*** -0.042*** -0.040*** -0.036***

(0.000) (0.000) (0.000) (0.000)

FREVt 1.423*** 1.303*** 1.600*** 1.272***

(0.000) (0.000) (0.000) (0.000)

SUEt 0.000* 0.000* 0.000 0.000*

(0.079) (0.078) (0.339) (0.089)

SGt 0.248*** 0.252*** 0.287*** 0.249***

(0.000) (0.000) (0.000) (0.000)

TAt 0.181** 0.189** 0.211** 0.237***

(0.015) (0.011) (0.013) (0.003)

CAPEXt 0.425*** 0.374*** 0.271** 0.412***

(0.000) (0.000) (0.014) (0.000) BPt -0.359*** -0.361*** -0.303*** -0.333***

(0.000) (0.000) (0.000) (0.000) EPt 0.143* 0.160** 0.174* 0.074

(0.087) (0.035) (0.064) (0.364)

Adjusted R2 0.086 0.087 0.083 0.084

Observations 10,286 10,286 7,044 10,226

27

TABLE 5 Regressions of FERC Models

Panel A: Main Tests Managerial ability (ABILITY) proxies =

Press Citation Award DEA Score IndAdjROA

Variables

Coefficient estimate

Coefficient estimate

Coefficient estimate

Coefficient estimate

Intercept -0.227*** -0.217*** -0.377*** -0.276***

(0.000) (0.000) (0.000) (0.000) Xt-1 -0.744*** -0.740*** -0.772*** -0.854***

(0.000) (0.000) (0.000) (0.000) Xt 0.349*** 0.530*** 0.557*** 0.549***

(0.000) (0.000) (0.000) (0.000) Xt3 1.032*** 1.001*** 1.055*** 0.757***

(0.000) (0.000) (0.000) (0.000) Rt3 -0.104*** -0.107*** -0.095*** -0.106***

(0.000) (0.000) (0.000) (0.000) ABILITYt -0.002*** -0.046** 0.039 0.201***

(0.004) (0.043) (0.387) (0.000) ABILITYt * Xt-1 0.008 0.144 -0.748 -0.701**

(0.317) (0.649) (0.209) (0.030) ABILITYt * Xt 0.010* -0.035 0.487 1.994***

(0.162) (0.903) (0.434) (0.003) ABILITYt * Xt3 0.007* 0.249** 0.470** 0.563*

(0.094) (0.048) (0.046) (0.057) ABILITYt * Rt3 0.001*** 0.049*** 0.095** 0.148***

(0.009) (0.003) (0.018) (0.002) RECt 0.088*** 0.094*** 0.097*** 0.094*** (0.000) (0.000) (0.000) (0.000) RECt * Xt3 0.128*** 0.096*** 0.136*** 0.097*** (0.001) (0.006) (0.004) (0.007) SIZEt -0.044*** -0.045*** -0.036*** -0.045***

(0.000) (0.000) (0.000) (0.000) SIZEt * Xt3 0.010 0.005 -0.020 0.043**

(0.652) (0.806) (0.488) (0.035)

LOSSt -0.036** -0.021 -0.047** -0.020

(0.035) (0.202) (0.019) (0.248)

LOSSt * Xt3 -0.253*** -0.276*** -0.115* -0.161***

(0.000) (0.000) (0.078) (0.002)

BMt 0.099*** 0.114*** 0.159*** 0.153***

(0.000) (0.000) (0.000) (0.000)

BMt * Xt3 -0.112** -0.130** -0.025 -0.102*

(0.050) (0.017) (0.688) (0.070)

EARNSTDt 0.630*** 0.600*** 0.703*** 0.650***

28

(0.000) (0.000) (0.000) (0.000)

EARNSTDt * Xt3 -0.987*** -0.571*** -0.765*** -0.715***

(0.000) (0.000) (0.000) (0.000)

NANALt 0.125*** 0.113*** 0.140*** 0.122***

(0.000) (0.000) (0.000) (0.000)

NANALt * Xt3 -0.198*** -0.144*** -0.114* -0.180***

(0.000) (0.002) (0.068) (0.000)

Adjusted R2 0.186 0.195 0.210 0.188

Observations 10,286 10,286 7,044 10,226